JPH1041205A - Scanning alligner - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce misalignment between a reticle and a wafer without exciting mechanical resonance, etc., of a stage system during scanning exposure. SOLUTION: Four positions WX1, WX2, WY1 and WY2 on a wafer stage which are measured with an interferometer 17 are supplied to a calculating means 15c. A calculation means 15c and differential devices 31X-31R found differences ΔRX, ΔRL and ΔRR between the target reticle stage position which is found based on the four positions, and the actual reticle stage position, and the differences are fed back to a reticle stage control system 17. Further, a correction speed αX.VW* corresponding to a tilt angle between the wafer stage and the reticle stage in the scanning direction is found with a calculating means 15d, and the correction speed is supplied to the reticle stage control system 17 by a feed forward method, for correcting the inclination angle.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for transferring a mask pattern onto a photosensitive substrate during a photolithography process for manufacturing, for example, a semiconductor device, an image pickup device (such as a CCD), a liquid crystal display device, or a thin film magnetic head. More specifically, the present invention relates to a scanning type exposure apparatus such as a step-and-scan type projection exposure apparatus for performing transfer by synchronously scanning a mask and a photosensitive substrate with respect to a projection optical system. .

[0002]

2. Description of the Related Art For example, when manufacturing a semiconductor device, a step-and-repeat method has conventionally been used as an exposure apparatus for transferring a reticle pattern as a mask to each shot area of a wafer coated with a photoresist. A reduction projection type exposure apparatus (stepper) of a batch exposure type has been frequently used. On the other hand, recently, in order to respond to the demand for transferring a large-area circuit pattern with high accuracy without significantly increasing the burden on the projection optical system, a part of the pattern on the reticle is transferred through the projection optical system. A so-called step-and-scan method in which the image of the pattern on the reticle is sequentially transferred to each shot area on the wafer by synchronously scanning the reticle and the wafer with respect to the projection optical system while being projected on the wafer. Projection exposure apparatuses have been developed.

A conventionally known aligner that transfers a pattern on the entire surface of a reticle onto the entire surface of a wafer with an equal-magnification normal image in a single scanning exposure using an integrated stage can be said to be a prototype of a scanning exposure apparatus. It is. On the other hand, in the step-and-scan method, since a projection optical system with a normal reduction magnification is used, it is necessary to drive the reticle stage and the wafer stage independently at a speed ratio according to the reduction magnification, and Since the movement between the shot areas is performed by the stepping method, the mechanism of the stage system is complicated and requires extremely sophisticated control.

A control system of a stage system of a conventional step-and-scan type projection exposure apparatus is as follows, for example. That is, after the alignment between the reticle stage and the wafer stage is performed before the start of scanning, the reticle is moved at a speed corresponding to the predetermined speed and in a corresponding direction in synchronization with scanning the wafer stage at a predetermined speed in a predetermined direction. While scanning the stages, the laser interferometer measures the displacement of both stages in the scanning direction and the non-scanning direction (the direction perpendicular to the scanning direction), and the thus measured displacement is servo-controlled. To make it smaller. In other words, conventionally, the amount of misalignment measured in the closed loop is fed back to reduce the amount of misalignment.

[0005]

As described above, in the conventional step-and-scan type projection exposure apparatus, the position shift amount is fed back by a servo control method at the time of scanning exposure, so that the reticle stage and the wafer stage can be moved. The amount of displacement in the scanning direction and the non-scanning direction has been reduced. However, there is a predetermined limit to the response speed of a feedback system that operates following such a positional deviation amount, and therefore, there is a disadvantage that a certain degree of tracking error (positional deviation amount) usually remains.

Specifically, for example, the scanning directions of the wafer stage and the reticle stage should be parallel, but there may be a predetermined angle error between the scanning directions due to a mechanical assembly error or the like. . If such an angular error exists, a displacement amount in the non-scanning direction is always generated on the reticle stage side during the scanning exposure. Therefore, the displacement amount is fed back to correct the displacement amount. Power is generated. However, the response speed of the feedback system is limited,
A certain degree of tracking error always remains in the non-scanning direction.

In this regard, as one of the techniques for reducing such a tracking error, there is a technique for increasing the gain of the feedback system. However, if the gain of the feedback system is simply increased, a mechanical resonance phenomenon of the stage system is likely to be excited, and the stage system may vibrate during scanning exposure. In view of the foregoing, an object of the present invention is to provide a scanning exposure apparatus that can reduce the amount of positional deviation between a reticle and a wafer without exciting mechanical resonance or the like of a stage system during scanning exposure. I do.

[0008]

A scanning exposure apparatus according to the present invention provides a mask (12) on which a pattern for transfer is formed.
A mask-side stage (9 to 11) for moving the substrate, and a substrate-side stage (1 to 4) for moving the photosensitive substrate (5), and the mask (12) is illuminated with illumination light for exposure. Then, the mask (1) is passed through the mask-side stage (9 to 11).
2) The substrate (5) is scanned in a direction corresponding to the predetermined direction via the substrate-side stages (1 to 4) in synchronization with the scanning of the substrate (5) in the predetermined direction. In the scanning exposure apparatus for sequentially exposing the pattern of (12), arithmetic control means (15c, 31X) for generating control information for controlling the operation of the mask side stage and the substrate side stage while maintaining a predetermined positional relationship. , 32X)
Error correction for adding correction information for correcting a previously measured error in the positional relationship between the mask-side stage and the substrate-side stage to the control information generated by the operation control means (15c, 31X). Means (15d, 33);
It has.

According to the present invention, for example, as shown in FIG. 2, the positions and rotation angles of the mask-side stages (9 to 11) are measured at two measurement points arranged at predetermined intervals in the non-scanning direction. Are indicated by the position (RR, RL) in the scanning direction and the position (RX) in the non-scanning direction.
The position and the rotation angle of 4) are represented by the positions (WY1 and WY2) in the scanning direction at two measurement points and the positions (WX1 and WX2) in the non-scanning direction at the two measurement points. In this case, the positions (RR, RL, RX, WY1, WY2, W
X1, WX2) are measured with high accuracy by, for example, a laser interferometer.

Then, before starting the scanning, the mask (12)
In a state where the positioning between the mask-side stage and the substrate-side stage is performed in a state where the positioning between the mask-side stage and the substrate-side stage has been performed, the arithmetic control means (15c, 31)
X, 32X), for example, by performing a predetermined calculation on the positions (WY1, WY2, WX1, WX2) of the substrate-side stages (1 to 4), the corresponding mask-side stages (9 to 1).
The target position (RR ^* , RL ^* , RX ^* ) of 1) is obtained. Furthermore,
The arithmetic control means (15c, 31X, 32X) subtracts the actual position (RR, RL, RX) of the mask-side stage from the target position (RR ^* , RL ^* , RX ^* ) to obtain a positional shift as control information. The amount (ΔRR, ΔRL, ΔRX) is obtained and fed back, and the positional deviation amount is, for example, (0,
(0, 0), the position of the mask-side stage is corrected.

Further, in the present invention, the error of the positional relationship between the mask-side stage and the substrate-side stage is measured in advance. Then, the error correction means (15d, 33) obtains correction information for correcting the positional error, and adds the correction information to the control information generated by the arithmetic control means (15c, 31X). Based on the result of the addition, the position of the mask-side stage is corrected. This means that the mask-side stage is driven in a feedforward manner based on the correction information. As a result, without increasing the gain of the former feedback system, that is, without exciting mechanical resonance or the like of the stage system, an error in the positional relationship between the mask-side stage and the substrate-side stage, and eventually the mask (12). ) And the substrate (5) are less displaced.

In this case, one example of the error of the positional relationship measured in advance is a mechanical assembly error. This mechanical assembly error is, for example, an angular error between the scanning direction of the mask-side stage and the scanning direction of the substrate-side stage, or yawing or the like that occurs in any of the stages according to the scanning position. Since such a mechanical assembly error can be measured with high accuracy in advance, it can be corrected at high speed by the feedforward method according to the present invention.

Further, one of the mask-side stage and the substrate-side stage is composed of a coarse movement stage (9, 10) for scanning and a fine movement stage (11) for position correction. When an error in a certain positional relationship is an angle error in the moving direction of the coarse movement stage (9, 10), the arithmetic control means (15c, 31X, 32X) uses the speed information (11c) of the fine movement stage (11) as the control information. G
PX · ΔRX) and the error correction means (15d, 3
3) calculation and control means (15c, 3c) for correcting speed information (αX · VW) for correcting the angle error of the coarse movement stage.
1X, 32X), it is desirable to add to the speed information (GPX · ΔRX) of the fine movement stage.

In this case, for example, the coarse movement stage (9, 1)
0) is driven at a constant speed, and the fine movement stage (11) is driven so as to correct the remaining displacement. Since the fine movement stage (11) can be reduced in weight, the displacement amount can be corrected at a high response speed. Further, for example, as shown in FIG. 4, when there is an angular error φ between the scanning direction (23) of the substrate-side stages (1 to 4) and the scanning direction (24) of the coarse movement stages (9, 10). The substrate side stage (1 to
The amount of displacement in the non-scanning direction between 4) and the coarse movement stage (9, 10) gradually increases. Then, since the amount of displacement calculated by the arithmetic control means (15c, 31X, 32X) also increases, the speed (G
The fine movement stage (11) is driven in the non-scanning direction by (PX · ΔRX). However, since this operation is a feedback method, a certain error remains due to the limit of the response speed as it is.

Therefore, the error correction means (15d, 33) at a speed (αX · VW) corresponding to the angle error φ,
By driving the fine movement stage (11) in the non-scanning direction by the feed forward method, the angle error φ is corrected substantially in real time. Further, since the fine movement stage (11) is controlled by the speed, the angle error φ
Can be corrected simply by driving the fine movement stage (11) in the non-scanning direction at a constant speed, and the control is simple.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a scanning exposure apparatus according to the present invention will be described below with reference to the drawings.
In this example, the present invention is applied to a step-and-scan type projection exposure apparatus. FIG. 1 shows a projection exposure apparatus according to the present embodiment. In FIG. 1, a rectangular illumination area (hereinafter, referred to as a “slit illumination area”) 21R with illumination light IL for exposure from an illumination optical system (not shown) is shown. Illuminates the pattern on the pattern forming surface of the reticle 12, and the pattern in the illumination area 21R is reduced via the projection optical system 8 at a magnification β (β is, for example, ４ ，, ５, etc.) and the wafer The projection exposure is performed on the slit-shaped exposure area 21W on the upper surface 5 of FIG. Hereinafter, a description will be given by taking the Z axis in a direction parallel to the optical axis of the projection optical system 8, taking the X axis parallel to the plane of FIG. 1 and the Y axis perpendicular to the plane of FIG. 1 in a plane perpendicular to the Z axis. I do. When exposure is performed by the scanning exposure method, the illumination region 21R
Respect, the reticle 12 is synchronized to be scanned in the -Y direction (or the + Y direction) in velocity V _R, the wafer 5 is scanned in the + Y direction (or the -Y direction) in velocity beta · V _R .

The drive system for the reticle 12 and the wafer 5 will be described. A reticle Y-axis drive stage 10 driven in the Y direction is mounted on a reticle support 9 and a reticle minute drive stage is mounted on the reticle Y-axis drive stage 10. 1
The reticle 12 is held on a reticle minute drive stage 11 by vacuum suction. The reticle minute drive stage 11 is a reticle 12 with a small amount and high precision in each of the X direction, the Y direction, and the rotation direction (θ direction) parallel to the plane of FIG. 1 in a plane perpendicular to the optical axis of the projection optical system 8. Is performed. Reticle micro drive stage 11
A movable mirror 13 is arranged on the upper side, and the positions of the reticle minute drive stage 11 in the X, Y and θ directions are constantly monitored by a reticle side interferometer 14 arranged on the reticle support 9. The position information S1 obtained by the interferometer 14 is supplied to the main control system 15.

On the other hand, a wafer Y-axis drive stage 2 driven in the Y direction is mounted on the wafer support table 1, and a wafer X-axis drive stage 3 driven in the X direction is mounted thereon. A Zθ axis driving stage 4 is provided thereon, and a wafer 5 is held on the Zθ axis driving stage 4 by vacuum suction. Moving mirror 6 on Zθ axis drive stage 4
Is fixed, the position of the Zθ-axis driving stage 4 in the X, Y and θ directions is monitored by the wafer-side interferometer 7 disposed outside, and the position information obtained by the interferometer 7 is also used as the main control system. 15. The main control system 15
The operations of the wafer Y-axis drive stage 2 to the Zθ-axis drive stage 4 are controlled via a wafer stage control system 16, and the operations of the reticle Y-axis drive stage 10 and the reticle minute drive stage 11 are controlled via a reticle stage control system 17. In addition to controlling, the operation of the entire apparatus is controlled in a general manner.

A reference mark member (not shown) on which a reference mark serving as a reference for alignment of the reticle 12 and the wafer 5 is fixed on the Zθ-axis drive stage 4, and above the reticle 12, A reticle alignment microscope (not shown) for detecting a positional shift amount between the reference mark and the mark on the reticle 12 is provided.
Further, an off-axis type alignment sensor 18 for detecting the position of a wafer mark for alignment on the wafer 5 is disposed on a side surface of the projection optical system 8 in the Y direction. The alignment between the reticle 12 and the wafer 5 is performed by these reference mark members, the alignment sensor 18 and the like.

Next, a detailed configuration of the wafer stage and the reticle stage will be described with reference to FIG. FIG.
2A is a plan view of the wafer stage, and FIG.
3A, the wafer 5 is placed on the Zθ axis driving stage 4.
Is placed. An X-axis movable mirror 6X and a Y-axis movable mirror 6Y are fixed on the Zθ axis drive stage 4 so that a reticle pattern image is projected onto a slit-like exposure area 21W on the wafer 5 during exposure. Has become. The Zθ-axis driving stage 4 is driven in the Y direction (scanning direction) by the linear motor system by the wafer Y-axis driving stage 2 in FIG. 1 and is driven in the X direction by the wafer X-axis driving stage 3 by the feed screw system. It is supposed to be.

The movable mirror 6X has an X-axis interferometer 7X.
1 and 7X2, the laser beams 19X1 and 19X2 are spaced at intervals L along an optical path parallel to the X axis and passing through the optical axis of the projection optical system 8 and the optical axis of the alignment sensor 18, respectively.
The movable mirror 6Y is irradiated with laser beams 19Y1 and 19Y2 from the interferometers 7Y1 and 7Y2 for the Y axis at intervals L along an optical path parallel to the Y axis. In this example, the X-axis measurement values WX1 and WX2 by the interferometers 7X1 and 7X2 are used as the position information of the Zθ-axis drive stage 4.
And the measured value WY1 of the Y axis by the interferometers 7Y1 and 7Y2
And four-degree-of-freedom information composed of WY2 and WY2. Therefore, in this example, in addition to the position of the Zθ-axis drive stage 4 in the X and Y directions, the rotation angle due to yawing can be detected. For example, the rotation angle can be detected even if the interferometer 7X2 or the interferometer 7Y2 is omitted. However, as in this example, the position and yawing are obtained by detecting the position information having four degrees of freedom and obtaining the averaging effect. Detection accuracy is increasing.

When the off-axis alignment sensor 18 is used, the position in the X direction is controlled based on the measurement value of the interferometer 7X2 using the laser beam 19X2. As a result, a so-called Abbe error does not occur. FIG. 2B is a plan view of the reticle stage. In FIG. 2B, a reticle minute drive stage 11 is mounted on a reticle Y-axis drive stage 10 and a reticle 12 is held thereon. I have. The reticle minute drive stage 11 has a movable mirror 13X for the X axis and two corner cubes 13L and 13 for the Y axis.
R is fixed, and the movable mirror 13X is irradiated with the laser beam 20X from the interferometer 14X in parallel with the X axis. Also,
The interferometers 14L and 14R for the Y axis are spaced from the cube cubes 13L and 13R by a distance M parallel to the Y axis.
Are irradiated with the laser beams 20L and 20R.

The laser beams 20L and 20R reflected by the corner cubes 13L and 13R are respectively reflected by the corner cubes 13L and 13R via the reflection mirrors 22L and 22R.
After being returned to 13R, it is returned to interferometers 14L and 14R. That is, the reticle interferometers 14L and 14R are double-pass interferometers, so that the laser beam does not shift due to the rotation of the reticle minute drive stage 11. Further, at the time of exposure, the slit-shaped illumination region 21R on the reticle 12 is illuminated by the exposure light with uniform illuminance.

The reticle Y-axis drive stage 10 of the present embodiment
The reticle minute drive stage 11 is driven by a linear motor system in the Y direction along two drive axes substantially parallel to the laser beams 20L and 20R, and the reticle Y-axis drive stage 10 is substantially parallel to the laser beam 20X. It is driven by an actuator (not shown) in the X direction along the drive shaft. In this case, by changing the balance between the drive amounts of the two Y-axis linear motors, the reticle Y-axis drive stage 10 and, consequently, the reticle minute drive stage 11 can be rotated within a predetermined range. That is, the reticle minute drive stage 11 is driven by a drive system having three degrees of freedom as a whole.

As the position information of the reticle minute drive stage 11 in the X direction, a measurement value RX by the interferometer 14X is used, and the position of the reticle minute drive stage 11 in the Y direction along the drive axes of the two linear motors is used. As the position information, the measurement values RL and R measured by the interferometers 14L and 14R are used.
R is to be used. As can be seen from FIGS. 1 and 2, the reticle-side interferometer 14 in FIG. 1 generically includes three interferometers 14X, 14R, and 14L, and the wafer-side interferometer 7 includes four interferometers 7X1, 7X2,7Y1,7
Y2 is a general term.

The reticle micro-drive stage 11 can be finely moved with three degrees of freedom along a drive axis substantially parallel to the laser beams 20L, 20R, 20X with respect to the reticle Y-axis drive stage 10. You may. In this case, the stage on the reticle side is driven with five degrees of freedom as a whole. For example, two linear motors for driving the reticle Y-axis drive stage 10 in the Y direction are driven at a constant speed in an open loop. By controlling the drive system having three degrees of freedom for finely moving the reticle minute drive stage 11 in a closed loop to correct the displacement between the reticle 12 and the wafer 5, the control system of the present example can be used almost as it is.

Next, a synchronous control method of the stage system during scanning exposure of the projection exposure apparatus of this embodiment will be described with reference to FIG. FIG. 3 shows the configuration of the control system of the stage system of this example. In FIG. 3, it is assumed that alignment between the reticle and the wafer is performed, and the reticle and the wafer are respectively positioned at the scanning start positions. And the main control system 1
The control means 15a in 5 transmits a command for starting scanning exposure to the speed command generating means 15b, the calculating means 15c, and the multiplying means 1.
5d, the speed command generating means 15b sends a signal to the wafer stage control system 16 and the multiplying means 15d as shown in FIG.
(Usually 0) of the target speed of the wafer side in the X direction Zθ axis driving stage 4, and the speed command signal indicating the target scanning speed V _W ^* of the Y direction is supplied. The latter target scanning speed V _W ^{* in} the Y direction gradually increases from 0 to a constant scanning speed V _W ^*.
After a lapse of a predetermined time after reaching _W0 , the shape gradually changes to a trapezoidal shape with respect to time, such as gradually returning to 0 again. The control means 15a, the speed command generation means 15b, the calculation means 15
c and the multiplication means 15d are functions on software of a computer.

In response, the multiplying means 15d multiplies the target scanning speed V _W ^* by a coefficient αX which is predetermined and stored as described later, and calculates a correction speed αX · V _W ^* obtained as a result. Adder 3 in reticle stage control system 17
3 to one input. Further, the wafer stage control system 16 drives the wafer Y-axis drive stage 2 of FIG. 1 in the Y direction at its target scanning speed _VW ^* by open loop control, and moves the wafer X-axis drive stage 3 to the target speed if necessary. To drive in the X direction. At this time, Zθ measured by the four interferometers 7X1, 7X2, 7Y1, 7Y2 in FIG.
Positions WX1, WX2, WY1, WY of shaft drive stage 4
2 is the arithmetic means 15 of FIG.
c. In the calculating means 15c, first, FIG.
Using the distance L between the two laser beams 19X1 and 19X2 and the distance L between the laser beams 19Y1 and 19Y2,
A vector composed of the measured values WX1, WX2, WY1, WY2 is multiplied by a matrix T1 defined by the following equation, and is composed of a position WX in the X direction of the Zθ axis drive stage 4, a position WY in the Y direction, and a rotation angle Wθ. Convert to vector.

[0029]

(Equation 1)

Next, the arithmetic means 15c calculates the relative rotation angle θ between the reticle 12 and the wafer 5 determined by the alignment and the function of the projection magnification β from the reticle 12 to the wafer 5 as the following equation. Introduce matrices T3, T2 and offset vector B. Then, a vector (W) having the position and the rotation angle of the Zθ axis driving stage 4 as components.
(X, WY, Wθ), a vector (RX ^* ) having the target position RX ^{* in} the X direction, the target position RY ^{* in} the Y direction, and the target rotation angle Rθ ^* of the reticle minute drive stage 11 in FIG. , RY ^* , Rθ ^* ).

[0031]

(Equation 2)

Further, the calculation means 15c uses the matrix T4 having the interval M between the laser beams 20L and 20R of FIG. 2 (b) as a component as shown in the following equation to obtain the vectors (RX ^* , RY ^* , Rθ).
^* ) Is a vector (RX ^* , vector) consisting of target positions corresponding to the three drive axes of the reticle minute drive stage 11 shown in FIG.
RL ^* , RR ^* ). In other words, the vector (R
X ^* , RL ^* , RR ^* ) are the interferometers 14X, 14X during scanning exposure.
A vector indicating a target position of the reticle minute drive stage 11 to be measured by L, 14R. Each of the elements RX ^* , RL ^* , RR ^* of the vector is
X, 31L, and 31R are supplied to the input section on the addition side.

[0033]

(Equation 3)

In FIG. 3, the difference units 31X, 31
The positions RX, RL, RR actually measured by the reticle-side interferometers 14X, 14L, 14R are also supplied to the subtraction-side inputs of L, 31R, respectively. Differentiator 3
In 1X, 31L, and 31R, as a whole, a vector (RX ^* , RL ^* , RR ^* ) indicating the target position of the reticle minute drive stage 11 is calculated from a vector (RX ^* , RL ^* , RR ^* ) indicating the actual position of the reticle minute drive stage 11 as RX, RL, RR)
Is subtracted to obtain an error vector (ΔRX, ΔRL, ΔRR)
Ask for. This error vector is a vector (WX1, WX2, WY) indicating the position of the Zθ axis driving stage 4 described above.
1, WY2), and matrices T1 to T4.

[0035]

(Equation 4)

Then, the elements ΔRX, ΔRL, ΔΔ of the error vectors are obtained from the difference units 31X, 31L, 31R, respectively.
The RR is supplied (feedback) to the reticle stage control system 17, and the reticle stage control system 17 performs the reticle minute drive shown in FIG. 2B so that the elements ΔRX, ΔRL, and ΔRR of the supplied error vector become 0, respectively. The drive system of the stage 11 having three degrees of freedom is controlled.

In this embodiment, the speed of the drive system of the reticle minute drive stage 11 is controlled by converting the element of the error vector into a speed. That is, in the reticle stage control system 17, the elements ΔRX, ΔRL,
ΔRR is a position gain unit 32X, 32L, 32R, respectively.
, And are multiplied by the position gains GPX, GPL, and GPR by the position gain units 32X, 32L, and 32R, respectively, and are converted into speeds. The speed GPX · ΔRX output from the position gain unit 32X is supplied to the other input unit of the adder 33, and the adder 33 corrects the corrected speed αX · V _W ^* from the multiplication unit 15d and its speed GPX · ΔRX. Is supplied to the speed gain unit 34X. The speed supplied by the speed gain unit 34X is converted into drive power, and the drive power drives the motor 35X for the X-axis actuator.
1 slightly moves in the X direction. In addition, actually, the speed gain unit 3
The measured value of the speed of the motor 35X is fed back just before the input section of 4X, but its feedback circuit is omitted. The same applies to other motors.

In parallel with this, the position gain units 32L and 32L
Speeds GPL · ΔRL and GPR · Δ output from 2R
RR is supplied to speed gain units 34L and 34R, respectively, and the reticle Y-axis drive stage 1 is driven by drive power output from speed gain units 34L and 34R, respectively.
The linear motors 35L and 35R for driving 0 (and, consequently, the reticle minute drive stage 11) in the Y direction and the rotation direction are driven. And those motors 35X,
The position of the reticle micro-drive stage 11 driven by the linear motors 35L, 35R is shifted by three interferometers 14X,
As the measurement values of 14L and 14R, the difference units 31X, 31L,
The reticle minute drive stage 11 is fed back to 31R and driven by closed loop control. As a result, when the movement directions of wafer Y-axis drive stage 2 and reticle Y-axis drive stage 10 are parallel, Zθ-axis drive stage 4 (wafer 5) is driven at a constant speed V _W0 in the + Y direction. Reticle minute drive stage 1 in synchronization with
1 (reticle 12) is driven in the −Y direction at a speed (1 / β) V _W0 while maintaining the relative rotation angle θ.

However, in practice, it is assumed that the moving direction of the wafer Y-axis driving stage 2 in FIG. 1 is parallel to the Y-axis.
For example, reticle Y-axis drive stage 1 due to an assembly error
The movement direction of 0 may be inclined with respect to the Y axis.
In such a case, simply feeding back the error vector leaves a predetermined displacement amount.
In this example, the position shift amount is corrected by the feedforward method using the multiplication means 15d and the adder 32X of FIG.

FIGS. 4A and 4B are plan views of the stage on the reticle side. In FIGS. 4A and 4B, the projection of the Zθ axis drive stage 4 on the wafer side onto the reticle side is performed. The image is shown as a dotted projection image 4R. At the time of scanning exposure, the reticle Y-axis drive stage 10 is tilted from the + Y direction by a predetermined minute tilt angle φ in synchronization with the projection image 4R of the wafer-side Zθ-axis drive stage 4 being driven, for example, in the + Y direction. Driven in the direction. That is, the velocity vector 23 of the projection image 4R is oriented in the + Y direction, and the velocity vector 24 of the reticle Y-axis drive stage 10 is inclined at an inclination angle φ with respect to the + Y direction.

Here, as shown in FIG. 4A, at a certain point, a projected image 25 of the circuit pattern on the wafer onto the reticle is obtained.
FIG. 4 shows a state in which the reticle minute drive stage 11 has moved by the distance LY in the direction of the velocity vector 24 from that time on assuming that the pattern 26 and the pattern 26 on the reticle overlap.
It is assumed that the state is as shown in FIG. At this time, the projected image 4R is also +
It has moved almost LY in the Y direction. Then, when the position adjustment of the reticle minute drive stage 11 in the X direction is not performed, in other words, when the feedback of the error vector element ΔRX and the feedforward of the correction speed αX · V _W ^* are not performed in FIG. FIG. 4 (b)
As shown in FIG. 7, between the projected image 25 of the circuit pattern on the wafer and the pattern 26 on the reticle, almost LY · sin
A positional shift amount in the X direction corresponding to φ (= ΔX) occurs.

At this time, the reticle minute drive stage 11
Assuming that the reflection surface of the movable mirror 13X (see FIG. 2B) is parallel to the Y axis, the position RX of the reticle micro-drive stage 11 measured via the laser beam 20X changes by the positional deviation amount ΔX. . Therefore, assuming that only feedback control of the element ΔRX of the error vector in FIG. 3 is performed, as shown in FIG. 4B, the driving force indicated by the speed vector VRX acts on the reticle minute driving stage 11, and The position of the reticle minute drive stage 11 is gradually displaced in the −X direction, and the displacement in the X direction caused by the inclination angle φ is corrected.

However, since there is a limit to the following speed of the feedback control, as shown in FIG. 5A, the speed vector 27 of the reticle minute drive stage 11 is reduced.
A is an intermediate vector between the speed vector 23 of the projection image 4R of the wafer-side Zθ axis drive stage 4 and the speed vector 24 of the reticle Y axis drive stage 10, and an error vector 28 remains. In this example, the error vector 28
Is supplied to the adder 33 of FIG. 3 in a feedforward manner in order to reduce the correction speed corresponding to the inclination angle φ. In this case, if the actual scanning speed of the Zθ axis driving stage 4 on the wafer side is set to the target scanning speed V _W ^* using the projection magnification β of the projection optical system 8, the reticle minute driving stage 11
Is approximately V _W ^* / β. Also, the position of the reticle minute drive stage 11 shifts in the + X direction by sin φ · V _W ^* / β per unit time.
It is only necessary to drive in the X direction at a speed of -sin φ · V _W ^* / β. Accordingly, the coefficient αX by which the target scanning speed V _W ^* is multiplied by the multiplying means 15d in FIG. 3 is as follows.

[0044]

ΑX = −sin φ / β Using this coefficient αX, the correction speed αX · V _W ^* calculated by the multiplication means 15d is supplied to the adder 33, so that the reticle minute drive is performed in a feed-forward manner. The stage 11 is designed to offset the amount of displacement due to the inclination angle φ.
Driven in the X direction. As a result, as shown in FIG. 5B, the actual speed vector 27B of the reticle minute drive stage 11 becomes substantially equal to the speed vector 23 of the projection image 4R of the wafer-side Zθ axis drive stage 4, and the reticle 12 Synchronous scanning is performed with the wafer and the wafer 5 aligned with high accuracy. At this time, since it is not necessary to increase the gain in the position gain devices 32X, 32L, 32R in the closed loop of FIG. 3, mechanical resonance of the stage system is not easily excited, and vibration during scanning exposure is reduced.

In this embodiment, since the operation of the reticle minute drive stage 11 is controlled by adding the correction speed in a feedforward manner, the inclination angle of the reticle Y-axis drive stage 10 in the moving direction is corrected. Only needs to add a constant correction speed αX · V _W ^*
Easy to control. In the above embodiment, the scanning direction of the wafer stage and the reticle Y-axis driving stage 10
Motor 3 for driving the reticle minute drive stage 11 in the X direction in order to correct the tilt angle with respect to the scanning direction.
The correction speed is added only on the 5X side by the feedforward method. However, for example, when it is known that yawing occurs at a predetermined position when the reticle Y-axis driving stage 10 scans in the Y direction, the correction speed for correcting the yawing is adjusted by the reticle. What is necessary is just to add to the linear motors 35L and 35R for driving the Y-axis drive stage 10 in the Y direction by a feedforward method. For this purpose, an adder may be inserted before the speed gain units 34L and 34R in FIG. 3 and the correction speed obtained by the predetermined arithmetic means may be supplied to these adders.

In the configuration of FIG. 3, the correction speed is added by the feedforward method. However, for example, by adding the correction position to the error vector elements .DELTA.RX, .DELTA.RL, and .DELTA.RR by the feedforward method, the position deviation is added. The amount may be corrected. Further, as described above, the reticle Y-axis driving stage 10 is driven at a constant speed, and the correction of the positional deviation amount is performed by the reticle minute driving stage 1.
1 may be performed. In this case, FIG.
As shown in the above, the target scanning speed V _W ^* / β on the reticle side is obtained from the target scanning speed V _W ^* from the speed command generating means 15b by the multiplying means 36, and the reticle Y axis is calculated using the target scanning speed V _W ^* / β What is necessary is just to drive the drive part 10a of the drive stage 10. Further, the wafer stage control system 16 may be controlled based on the position.

In the configuration of FIG. 1, the reticle 12 is mounted on the reticle Y-axis driving stage 10 via the reticle minute driving stage 11, but the reticle minute driving stage 11 is omitted and the wafer Y-axis driving Fine movement stages for changing the position of the wafer 5 within a predetermined range in the Y direction may be arranged at two positions separated by a predetermined distance in the X direction between the stage 2 and the wafer 5. In this case, the reticle Y-axis drive stage 10 is controlled in an open loop, and the wafer Y-axis drive stage 2 on the wafer side and the fine movement stage are controlled in accordance with the target position calculated from the measured position of the reticle Y-axis drive stage 10. And the like are controlled by a closed loop method.

Further, in the above example, as shown in FIG. 2, the position of the stage on the wafer side is measured with four degrees of freedom, and the position of the stage on the reticle side is measured with three degrees of freedom.
The positions of the stage on the wafer side and the stage on the reticle side may each be measured with three degrees of freedom. Alternatively, for example, the position of the stage on the wafer side may be measured with three degrees of freedom, and the position of the stage on the reticle side may be measured with four degrees of freedom. May be measured. As the degree of freedom of measurement increases, the averaging effect improves the position and rotation angle measurement accuracy.

As described above, the present invention is not limited to the above-described embodiment, and can take various configurations without departing from the gist of the present invention.

[0050]

According to the scanning exposure apparatus of the present invention,
There is provided an error correction means for adding correction information for correcting an error in a positional relationship between the mask-side stage and the substrate-side stage measured in advance to control information generated by the arithmetic control means. The error of the measured positional relationship can be corrected by the feedforward method. Therefore, there is no need to increase the gain when performing control by feeding back the control information generated by the arithmetic control means.
There is an advantage that the amount of misalignment between the mask and the substrate can be reduced without exciting mechanical resonance or the like of the stage system during scanning exposure.

If the previously measured positional relationship error is a mechanical assembly error, the positional relationship error is almost constant and can be accurately measured. There is an advantage that accurate correction can be made by the forward method. In addition, one of the mask-side stage and the substrate-side stage is composed of a coarse movement stage for scanning and a fine movement stage for position correction, and the error of the positional relationship measured in advance is the coarse movement stage. The arithmetic control means generates speed information of the fine movement stage as control information, and the error correction means calculates and controls corrected speed information for correcting the angle error of the coarse movement stage. In the case of adding to the speed information of the fine movement stage generated by the means, a simple control that only adds a constant correction speed corresponding to the angle error in a feed-forward manner is used, and the position shift due to the angle error is performed. There is an advantage that the amount can be corrected.

[Brief description of the drawings]

FIG. 1 is a front view mainly showing a stage mechanism of an example of an embodiment of a scanning exposure apparatus according to the present invention.

2A is a plan view showing a stage configuration on a wafer side in FIG. 1, and FIG. 2B is a plan view showing a stage configuration on a reticle side in FIG.

FIG. 3 is a configuration diagram including a partial functional block diagram showing a control system of a stage system according to the embodiment.

FIG. 4 is an explanatory diagram of a positional shift amount generated when there is an inclination angle φ between the scanning direction of the wafer stage and the scanning direction of the reticle Y-axis drive stage 10.

FIG. 5A shows a velocity vector 27A of the reticle minute drive stage 11 when only feedback control is performed.
(B) shows the velocity vector 27 of the reticle minute drive stage 11 when the feedforward control is also performed.
FIG.

[Explanation of symbols]

 Reference Signs List 1 wafer support table 2 wafer Y-axis drive stage 5 wafer 7X1, 7X2, 7Y1, 7Y2 wafer-side interferometer 8 projection optical system 9 reticle support table 10 reticle Y-axis drive stage 11 reticle minute drive stage 12 reticle 14X, 14L, 14R Reticle-side interferometer 15 Main control system 15b Speed command generating means 15c Computing means 15d Multiplying means 16 Wafer stage control system 17 Reticle stage control system 32X, 32L, 32R Position gain device 34X, 34L, 34R Speed gain device

Claims (3)

[Claims] 1. A mask-side stage for moving a mask on which a pattern for transfer is formed, and a substrate-side stage for moving a photosensitive substrate, wherein the mask is illuminated with illumination light for exposure. Scanning the substrate in a direction corresponding to the predetermined direction via the substrate-side stage in synchronization with scanning the mask in a predetermined direction via the mask-side stage. In a scanning exposure apparatus for sequentially exposing a pattern of a mask, an arithmetic control unit for generating control information for controlling operations of the mask-side stage and the substrate-side stage while maintaining a predetermined positional relationship, and the mask-side stage Error correction for adding correction information for correcting an error of a positional relationship measured in advance with the substrate-side stage to control information generated by the arithmetic control means. Scanning exposure apparatus characterized by having a means. 2. The scanning exposure apparatus according to claim 1, wherein the previously measured positional relationship error is a mechanical assembly error. 3. The scanning exposure apparatus according to claim 1, wherein one of the mask-side stage and the substrate-side stage comprises a coarse movement stage for scanning and a fine movement stage for position correction. The error of the positional relationship measured in advance is an angular error in the moving direction of the coarse movement stage, the arithmetic control means generates speed information of the fine movement stage as the control information, Means for adding correction speed information for correcting an angle error of the coarse movement stage to speed information of the fine movement stage generated by the arithmetic control means.